A Comparison of Allergen Sensitization Profiles in Patients with Chronic Rhinosinusitis with and Without Nasal Polyposis

Abstract

Chronic rhinosinusitis (CRS) and allergic rhinitis (AR) are common comorbid sinonasal conditions. CRS is classically divided into two distinct phenotypes: CRS with nasal polyposis (CRSwNP) and CRS without nasal polyposis (CRSsNP). The purpose of this retrospective observational study is to determine whether aeroallergen sensitization profiles in patients with comorbid CRS and AR can distinguish between CRSwNP and CRSsNP. A total of 241 patients diagnosed with comorbid CRS and AR who underwent skin prick testing or in vitro allergy testing in a single tertiary rhinology practice were included for evaluation. The rates of allergen-specific sensitizations in CRSwNP patients were compared with those in CRSsNP patients. Of the allergens tested in the routine panels, Dermatophagoides pteronyssinus (OR = 1.82, p = 0.03), Alternaria (OR = 2.55, p < 0.01), and animal dander (OR = 1.48 for cat and OR = 3.01 for dog, p < 0.01) were predictive of CRSwNP. Sensitization to any grass allergen was also predictive of CRSwNP (OR = 2.09, p < 0.01). Multiple perennial aeroallergens showed strong associations with CRSwNP; however, broad sensitization to perennial allergens as a whole group was not significantly predictive of CRSwNP (OR = 1.83, p = 0.22).

1. Introduction

Chronic rhinosinusitis (CRS) and allergic rhinitis (AR) are common sinonasal inflammatory conditions that affect more than 30 million people in the United States annually [1]. CRS is commonly divided into two phenotypes: chronic rhinosinusitis with nasal polyposis (CRSwNP) and chronic rhinosinusitis without nasal polyposis (CRSsNP). Both CRS and AR often manifest concomitantly with not only overlapping clinical symptomatology, but also similar immunologic pathophysiology that results in inflammation at the level of the sinonasal mucosa [2,3]. Although the exact pathways of CRSwNP pathogenesis are unclear, the general principles of functional breakdown in sinonasal mucosa and dysregulation of the immune response are understood to be associated with the generation of nasal polyps. Research studies suggest that disruptions to the protective sinonasal mucosal barrier through IL-4 and IL-6 decrease the expression of tight junctions, increasing tissue permeability [4,5]. Another mechanism of mucosal dysfunction includes decreased expression of anti-microbial proteins and increased expression of mucus-producing channels. Impairment of these sinonasal mucosal defense mechanisms increase exposure to microbes and pathogens that activate the immune system, creating chronic inflammation. The characterization of the inflammatory response in CRSwNP has been predominately associated with a type-2 or type-1 inflammatory response. Patients with histological evidence of type-2 inflammation present with elevated tissue eosinophils, IL-5, and mast-cells, while those with type-1 inflammation have elevated neutrophils and INF-γ [4,6]. Histopathological analysis of tissues allows for targeted biologic treatment, particularly in patients with type-2 inflammation. Biologics including omalizumab, mepolizumab, and dupilumab target IgE, IL-5, and IL-4, respectively. Targeting these specific inflammatory markers decreases polyp size and improves symptoms and can be used in conjunction with surgical intervention [7]. Through immunoglobulin E (IgE)-mediated reactions to inhaled allergen, AR is generally thought to exacerbate the severity of the CRS disease state by increasing the degree of mucosal inflammation within the sinonasal cavities and contributing to sinus ostial obstruction and mucus retention with the paranasal sinuses [8,9,10,11]. Specifically, perennial allergens, such as dust mites and molds, have been found to be associated with CRS more so than other aeroallergens; however, the exact role of atopy in the clinical manifestations of CRS has not been fully elucidated [12]. The objective of this study is to determine if, among patients with concomitant AR and CRS, specific allergen sensitization profiles can distinguish between CRSwNP and CRSsNP.

2. Materials and Methods

A retrospective observational study was conducted to evaluate adult patients with comorbid diagnoses of AR and CRS from a single tertiary rhinology practice between January 2014 and December 2020. The study protocol (18-06-WC-0163) was approved by the Institutional Review Board at Eastern Virginia Medical School prior to the initiation of the study (initial approval date 15 June 2018).

An applicable search of the electronic medical record was performed to identify subjects who were at least 18 years of age and demonstrated comorbid AR and CRS by querying according to International Classification of Disease, tenth revision (ICD-10), codes for AR (J30, J30.1, J30.2, J30.5, J30.8, J30.81, J30.89, and J30.9) and CRS (J32, J32.0, J32.1, J32.2, J32.3, J32.4, J32.8, and J32.9). Subjects were subsequently confirmed for inclusion in the retrospective analysis based upon additional chart review. Evidence of CRS in subjects were demonstrated by symptomatology based on the American Academy of Otolaryngology—Head and Neck Surgery CRS criteria and supporting endoscopic nasal examination and computed tomography (CT) findings. This study excluded subjects with secondary causes of CRS, including sinonasal neoplasm, iatrogenic injuries, and cystic fibrosis. AR was identified through a clinical history of seasonal or perennial allergic symptoms and previous documentation of either skin prick allergy testing or in vitro allergy testing. For subjects who had completed skin prick allergy testing, comprehensive results characterized sensitivities to a 38-aeroallergen skin prick panel. Positive results for the skin prick testing were defined as at least a 4 mm wheal in response to a specific aeroallergen exposure. Subjects who completed the in vitro allergy test received measurements of antigen-specific IgE levels to a set panel of inhalant allergens; positive results for in vitro allergy testing were defined by IgE > 0.35 kU.

For each subject, positive sensitization results to specific individual aeroallergens were collected from previous skin prick allergy testing and in vitro allergy testing. Allergen test results were analyzed with respect to sensitization to individual allergens, sensitization to a broader subgroup of allergen (tree, grass, mold, etc.), and in terms of total cumulative sensitizations. Additional data collected from the retrospective chart review included demographic data, including age, gender, ethnicity, and the presence of other comorbid atopic conditions, such as asthma, food allergies, and atopic dermatitis. The presence or absence of nasal polyposis, as seen on routine in-office nasal endoscopy, was used to separate the subjects into two distinct cohorts: CRSwNP and CRSsNP. The primary outcome measurement was the frequency of specific aeroallergen sensitizations for the CRSwNP and CRSsNP cohorts, allowing for comparison between the two distinct cohorts.

A power analysis was performed to determine the appropriate sample size for a study comparing allergen sensitivities in patients with CRSsNP and CRSwNP. The power analysis assumed a medium effect size (d = 0.5), a significance level of 0.05, and a desired power of 80%. Analysis indicated that approximately 124 patients per group were required to detect statistically significant differences in allergen sensitivities between the two groups using a two-sample t-test.

To test for violations of normalcy, univariate analyses were performed on all continuous variables. Pearson’s chi square test assessed associations between categorical patient descriptive characteristics, including allergen sensitizations and CRS phenotypes. Risk estimates and univariate regression were performed to assess individual relationships between significantly associated allergen sensitizations and CRS phenotypes. Predictors that were significant in univariate regression were tested with multivariate logistic regression. All hypothesis testing was carried out at the 95% significance level with p < 0.05 accepted as statistically significant. Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

3. Results

A total of 241 subjects were identified with comorbid diagnoses of AR and CRS. The mean age was 48 ± 15.4 years. Females constituted 59% of the subjects. Comorbid asthma, food allergies, and atopic dermatitis were diagnosed in 50%, 22%, and 7% of patients, respectively (Table 1). Aeroallergen sensitization was confirmed through skin-prick testing in 215 subjects and in vitro allergy testing in the remaining 26 subjects.

Table 1. Total Patient Demographics of 241 Subjects with Comorbid Chronic Rhinosinusitis and Allergic Rhinitis.

Demographic
Age (mean ± SD)	47.98 ± 15.37
Gender
Female	141 (59%)
Male	100 (41%)
Race
White	143 (59%)
Black	88 (37%)
Other	10 (4%)
Ethnicity
Hispanic	7 (3%)
Non-Hispanic	233 (97%)
Medical Comorbidities
Asthma	122 (50%)
Food Allergies	29 (12%)
Atopic Dermatitis	16 (7%)
Total Number of Sensitizations	12.4 ± 9.7

Among CRS phenotypes, 138 patients (57.3%) were defined as exhibiting CRSsNP, while 103 patients (42.7%) were classified as CRSwNP. Comparison between cohorts revealed that CRSwNP subjects were significantly younger than CRSsNP subjects (45.05 ± 15.28 years vs. 50.24 ± 2.46 years, p < 0.01). Asthma was more prevalent in CRSwNP than CRSsNP (OR = 2.8 (1.6–4.7 OR), p < 0.01). Across our study population, patients had an average of 12 unique allergen sensitizations; however, CRSwNP patients had an average of 15 which was significantly higher than the CRSsNP cohort with 11 (p < 0.01). Table 2 highlights the baseline characteristics of the two cohorts based on the presence or absence of nasal polyposis.

Table 2. Comparison of Demographic Variables for Subject Cohorts Differentiated by the Chronic Rhinosinusitis with and without Nasal Polyposis (CRSwNP and CRSsNP, respectively).

Demographics	CRSsNP (n = 138)	CRSwNP (n = 103)	p-Value
Age	50	45	0.01
Gender
Female	91 (66%)	50 (49%)
Male	47 (34%)	53 (51%)
Medical Comorbidities
Asthma	55 (39.9%)	67 (65%)	<0.01
Food Allergies	13 (9.4%)	17 (17%)
Atopic Dermatitis	12 (8.7%)	9 (22%)
Total Number of Sensitizations	10.6 ± 8.9	14.6 ± 10.2	<0.01

Aeroallergen sensitization profiles for the aggregated CRSsNP and CRSwNP cohorts were analyzed according to the following aeroallergen groupings: (1) epithelial allergens, including dust mites, animal dander, and cockroach (Table 3); (2) mold allergens (Table 4); (3) grass allergens (Table 5); (4) tree allergens (Table 6); and (5) weed allergens (Table 7). The aeroallergens were subsequently grouped as either seasonal or perennial. Seasonal aeroallergens were defined by grasses, trees, and weeds as a collective group, while perennial aeroallergens consisted of epithelial and mold allergens. Comparisons between CRSsNP and CRSwNP cohorts for different aeroallergen groupings are presented in Table 8.

Table 3. Comparison of Rates of Sensitization to Epithelial Allergens for CRSwNP and CRSsNP Cohorts.

Allergen	CRSwNP	CRSsNP	Odds Ratio	p-Value
Dermatophagoides pteronyssinus	63%	48%	1.82 (1.1, 3.1)	0.03
Dermatophagoides farinaemite	54%	42%	1.59 (0.9, 2.7)	0.09
Cat Dander	46%	28%	1.48 (1.17–3.8)	0.01
Dog Dander	28%	11%	3.01 (1.5, 6.0)	<0.01
American Cockroach	34%	30%	1.21 (0.7, 2.1)	0.5

Table 4. Comparison of Rates of Sensitization to Mold Allergens for CRSwNP and CRSsNP Cohorts.

Allergen	CRSwNP	CRSsNP	Odds Ratio	p-Value
Alternaria	48%	27%	2.55 (1.5, 4.4)	<0.01
Aspergillus	40%	23%	2.14 (1.2, 3.8)	0.01
Fusarium	37%	26%	1.67 (0.9, 3.0)	0.1
Helminthosporium	44%	30%	1.87 (1.1, 3.3)	0.03
Mucor	29%	20%	1.6 (0.83, 3.1)	0.16
Epicoccum	45%	25%	2.45 (1.4, 4.4)	<0.01
Neurospora	47%	28%	2.24 (1.2, 4.0)	0.01
Pullularia	39%	33%	1.34 (0.75, 2.40)	0.32
Rhizopus	25%	17%	1.69 (0.85, 3.4)	0.18

Table 5. Comparison of Rates of Sensitization to Grass Allergens for CRSwNP and CRSsNP Cohorts.

Allergen	CRSwNP	CRSsNP	Odds Ratio	p-Value
Timothy	41%	26%	2.00 (1.1, 3.5)	0.02
Bermuda	34%	33%	1.07 (0.6, 1.9)	0.81
Johnson	56%	30%	2.93 (1.7, 5.0)	<0.01
Bahia	53%	32%	2.43 (1.3, 4.6)	0.01

Table 6. Comparison of Rates of Sensitization to Tree Allergens for CRSwNP and CRSsNP Cohorts.

Allergen	CRSwNP	CRSsNP	Odds Ratio	p-Value
White Ash	37%	17%	2.81 (1.5, 5.3)	<0.01
White Alder	35%	25%	1.6 (0.9, 2.9)	0.14
American Beech	45%	24%	2.43 (1.3, 4.4)	<0.01
Red River Birch	53%	37%	1.92 (1.1,3.3)	0.02
Red Cedar	38%	21%	2.25 (1.2, 4.1)	0.01
Eastern Cottonwood	42%	26%	2.04 (1.1, 3.6)	0.02
American Elm	48%	33%	1.9 (1.1, 3.3)	0.02
Shagbark Hickory	56%	46%	1.50 (0.9, 2.6)	0.15
Box Elder	39%	29%	1.55 (0.9, 2.7)	0.12
White Pine	38%	26%	1.74 (1.0, 3.1)	0.07
White Oak	55%	36%	2.19 (1.3, 3.7)	<0.01
American Sycamore	48%	41%	1.37 (0.8, 2.4)	0.26
Black Walnut Pollen	45%	34%	1.54 (0.9, 2.7)	0.14
Black Willow Pollen	52%	39%	1.65 (0.9, 2.9)	0.08

Table 7. Comparison of Rates of Sensitization to Weed Allergens for CRSwNP and CRSsNP Cohorts.

Allergen	CRSwNP	CRSsNP	Odds Ratio	p-Value
Sheep Sorrel	50%	37%	1.67 (1.0, 2.8)	0.06
Lamb’s Quarter	41%	36%	1.22 (0.7, 2.1)	0.49
Rough Pigweed	31%	24%	1.44 (0.8, 2.6)	0.23
Ragweed	41%	27%	1.87 (1.1, 3.3)	0.03
English Plantain	49%	31%	2.19 (1.2, 3.8)	0.01
Mugwort Sage	53%	35%	2.08 (1.2, 3.6)	0.01

Table 8. Comparison of Sensitizations to Aeroallergen Groupings Across CRSwNP and CRSsNP Cohorts.

Allergen	CRSwNP	CRSsNP	Odds Ratio	p-Value
Grass	68%	50%	2.09 (1.22, 3.62)	0.007
Epithelial	90%	82%	1.87 (0.87, 4.29)	0.12
Trees	80%	71%	1.65 (0.87, 4.29)	0.11
Weeds	75%	64%	1.68 (0.95, 3.01)	0.07
Mold	73%	64%	1.48 (0.85, 2.62)	0.17
Seasonal	90%	82%	2.07 (0.97, 4.72)	0.06
Perennial	94%	90%	1.83 (0.71, 5.33)	0.22

3.1. Epithelial Allergens

Among the epithelial allergens that were tested in subjects, higher rates of sensitization to Dermatophagoides pteronyssinus, dog dander, and cat dander were associated with CRSwNP than CRSsNP (OR = 1.82, p = 0.03; OR = 3.01, p = 0.001; and OR = 1.48, p = 0.01, respectively). There was no significant association between sensitizations to Dermatophagoides farinaemite or American cockroach and a diagnosis of CRSsNP or CRSwNP. Sensitization to epithelial allergen as a whole group was not significantly associated with either CRSwNP or CRSsNP.

3.2. Mold Allergens

The CRSwNP cohort demonstrated higher (OR = 1.48, p = 0.17), though not statistically significant, rates of sensitization to mold aeroallergens as a whole. Individually, significant associations between sensitization and CRSwNP were identified for Alternaria, Aspergillus, Helminthosporium, Epicoccum, and Neurospora (OR = 2.55, p < 0.01; OR = 2.14, p = 0.01; OR = 1.87, p = 1.8; OR = 2.45, p < 0.01; OR = 2.24, p = 0.01, respectively).

3.3. Grass Allergens

Sensitizations to Timothy grass (OR = 2.00, p = 0.02), Johnson grass (OR = 2.93, p < 0.01), and Bahia grass (OR = 2.43, p = 0.01) were all more significantly associated with CRSwNP than with CRSsNP. In contrast, CRSsNP was associated with sensitization to Bermuda grass (OR = 1.07, p = 0.81). Generally, grass sensitization as a group was more greatly associated with CRSwNP (OR = 2.09, p = 0.007).

3.4. Tree Allergens

Sensitization to white ash (OR = 2.81, p = 0.001), American beech (OR = 2.43, p = 0.0003), red river beech (OR = 1.92, p = 0.02), red cedar (OR = 2.25, p = 0.01), eastern cotton wood (OR = 2.04, p = 0.02), American elm (OR = 1.9, p = 0.02), and white oak (OR = 2.19, p < 0.01) were all more significantly associated with CRSwNP than CRSsNP. Sensitization to tree allergens as a group was not significantly associated with presence or absence of nasal polyps.

3.5. Weed Allergens

Sensitization to ragweed (OR = 1.87, p = 0.01), English plantain (OR = 2.19, p = 0.01), and mugwort sage (OR = 2.08, p = 0.01) were all more significantly associated with CRSwNP than CRSsNP. Generally, however, sensitization to weed aeroallergens as a group was not associated with either CRSwNP or CRSsNP.

3.6. Seasonal Allergens Versus Perennial Allergens

Sensitization to seasonal allergens, consisting of grasses, trees, and weeds, was not found to significantly associated with either CRSwNP or CRSsNP. Likewise, sensitization to perennial allergens, consisting of dust mites, animal dander, cockroach, and molds, was not more significantly predictive of either CRSwNP or CRSsNP.

4. Discussion

Our findings demonstrate that sensitizations to specific molds, Dermatophagoides pteronyssinus, and animal dander were more strongly associated with CRSwNP. These results align with previous studies indicating that such allergens as molds may be associated with CRSwNP, as exhibited in patients with allergic fungal rhinosinusitis (AFRS). The pathophysiology behind AFRS is comparable to that of CRSwNP as disruption to the mucosal barrier with subsequent inflammatory response [13]. Interestingly, we observed that sensitization to D. Pteronyssinus was associated with CRSwNP, whereas D. farinae was not. This distinction suggests that allergic responses to dust mites and the development of nasal polyps are more nuanced in pathophysiology. Component-resolved diagnostic tests are a more nuanced, accurate allergy test that quantifies serum IgE levels targeted to specific proteins. Determining if a specific D. Pteronyssinus allergen protein is associated with CRSwNP would be an appropriate step towards understanding our findings [14]. Additionally, our study highlights the potential role of dog dander in the development of CRSwNP.

It is noteworthy that patients with CRSwNP exhibited higher rates of sensitization to individual allergens compared to CRSsNP. Despite this finding, broad sensitization across major allergen subgroups did not reach statistical significance (Table 8). Among subtypes, grass allergen was the only group for which sensitization was predictive of CRSwNP with 68% of CRSwNP patients demonstrating sensitization compared to 50% of CRSsNP patients. Further, the results of this study confirm the association between asthma and CRSwNP, as 65% of patients with CRSwNP were diagnosed with asthma [15].

Prior work by Odat et al. investigated the relationship between nasal polyps, asthma, and allergic rhinitis, specifically evaluating allergen sensitization patterns across CRS phenotypes. Their study, which assesses response to 24 unique allergens, found no significant difference in the rates of single allergen sensitivities between patients with CRSwNP and those with CRSsNP. Additionally, their study demonstrated that their patient population had an average of 2.6 positive allergens for the CRSsNP cohort and 2.5 positive allergens for the CRSwNP cohort [16]. Our findings varied significantly from those of Odat et al. In this study there was a significant difference between CRSsNP with CRSwNP cohort’s allergen sensitizations with 10.6 and 14.6 positive allergens, respectively. This difference suggests that individual allergen sensitizations play a role in the pathophysiology of CRSwNP and highlights a potential avenue of individualized allergen management strategies tailored to a specific patient’s profile.

This study was not designed to delineate whether the specific allergens or the chronic exposure to the allergens contributed more significantly to this difference. Future larger and geographically diverse studies would confirm these data regarding the relationship between allergen sensitizations. Further work to elucidate pathophysiology will help clarify the underlying associations presented in this study and guide clinical recommendations.

5. Limitations

Several limitations must be considered in the design of this study. The retrospective nature of the design allows for inherent bias that is difficult to control. Additionally, the patient demographics between the two groups did not match perfectly because this study was designed to capture the most random sample of patients from the investigators’ practice over the course of the specified time period. Additionally, the data from allergen tests completed at outside clinics did not use the same allergy extracts, potentially leading to confounding bias, but this source of bias should be prevalent among both groups and is a well-established flaw in skin prick allergy testing.

6. Conclusions

This study found that perennial allergens such as molds, Dermatophagoides pteronyssinus, and animal dander, were significantly associated with CRSwNP compared to CRSsNP. It is unclear whether the correlation is due to the irritant and reaction itself or due to the chronic exposure to the irritant. Perennial aeroallergens, among others, may provide potential targets for better understanding and treating CRSwNP in the future, and future work should focus on further delineating the nature of these associations.